Been noticing this for years honestly. There's a very consistent pattern where you get maybe 2-4 seconds of heavy static or what I'd describe as a kind of crackling white noise, and then the EVP comes through right after it. Not always clean but the timing is too regular to dismiss.
My working theory is that whatever is producing the anomalous audio needs some kind of electromagnetic baseline to work with, sort of like it's modulating on top of existing interference rather than generating sound from nothing. Which would explain why recordings in electrically active environments tend to yield more results.
What I'd want to know from others is whether the static matches on multiple devices running simultaneously. If you're getting teh same interference spike on two separate recorders at the exact same moment, that's a much stronger data point than a single device. Has anyone actually tested this properly with synced recordings?
This is the kind of thing that deserves more methodical attention than most EVP researchers give it.